oqp.analysis.descriptors

Excited-state character descriptors for MRSF states.

* ``participation_ratio`` -- NTO participation ratio PR = (sum w)^2 / sum w^2.
* ``tozer_lambda`` -- Tozer's Lambda: NTO-weighted spatial overlap of the
  hole/particle moduli (Lambda in [0,1]; high = local, low = charge transfer).
* ``fragment_ct_matrix`` -- Plasser/Lischka Omega matrix from the Loewdin-
  orthogonalized spin-flip transition density, partitioned over atom fragments.

Functions

participation_ratio(weights)
tozer_lambda(ao, nto_exc, grid_points, dV[, weight_thresh])
fragment_ct_matrix(states, ao, n, fragments)

Module Contents

participation_ratio(weights)

PR = (sum sigma^2)^2 / sum sigma^4 from NTO weights (= sigma^2).

tozer_lambda(ao, nto_exc, grid_points, dV, weight_thresh=1e-08)

Tozer Lambda for an excitation, from its hole/particle NTOs.

Lambda = sum_k w_k O_k / sum_k w_k, with O_k = \int |phi_hole,k||phi_part,k|.
The NTO orbitals are valence and smooth, so a moderate grid suffices.

fragment_ct_matrix(states, ao, n, fragments)

Fragment charge-transfer Omega matrix for excitation to root ``n``.

``fragments``: list of lists of 0-based atom indices.  Builds the AO
spin-flip transition density T = C_occ X^{(n)} C_vir^T, Loewdin-orthogonalizes
(Ttil = S^{1/2} T S^{1/2}), and sums Ttil_{mu,nu}^2 over fragment blocks.
Row index = hole fragment, column index = particle fragment.